The No.1 Website for Pro Audio
 Search This Thread  Search This Forum  Search Reviews  Search Gear Database  Search Gear for sale  Search Gearslutz Go Advanced
Tags: ,

Predicting Level Settings Without Reference? Effects Pedals, Units & Accessories
Old 2 weeks ago
  #1
Lives for gear
 

Thread Starter
Predicting Level Settings Without Reference?

How should location recording levels be set if the required levels are basically unknown, for example a situation where you can't test it in advance as it's mostly a unique event so you must be ready to record but have no idea about the sound pressure levels which will be reached:

YouTube

(a rare turbine trip test initiated by generator load shedding at near full power, while such tests are performed from time to time they're kept to the strict minimum because they highly stress the equipment).

Those interested in technical details can have a look at this video where it's dry:
YouTube

Logically dual recording seems appropriate but how many dB offset and there's still the question about how to set the "main" levels?

How about AGC and limiters? (Referring to location recording equipment.)
Old 2 weeks ago
  #2
Here for the gear
 

If you don't know the dB's at source, I don't recon there's a proper way to do it
Evaluate your distance from the sound source, take into account room reverberation, (you can ask someone to pop a ballon at the same distance and check those levels).
Loved the videos!
Interested in hearing what other guy's have to say about this.
Old 2 weeks ago
  #3
Gear Addict
 

That is where digital microphones proof their value. You can also think of using two preamps per microphone and set gains at 10 and 30. The most valuable tool is called experience. If you use very noise free converters you could set your gain at 10 or 20 dB, and add digital gain in post.
Old 2 weeks ago
  #4
Lives for gear
 

Thread Starter
Thanks for your replies.

For the shown situation it's very difficult to estimate required levels, as you've never seen it before you can only ask some employee who'll probably just tell you "It's loud" which won't help much.
Also there are no existing videos you could use as reference (for those interested, the video shows the upper surge tank chamber of the Mapragg power station in Switzerland during a simultaneous generator load shedding test of all 3 groups (total 330 MW, 345 MVA but the test is not performed at full load, here around 80-90 %), the load shedding initiates a turbine trip, in this example the wicket gate of each Francis turbine is closed (failsafe function) at the maximum allowable speed to prevent the runner from reaching its runaway speed, the pressure surge in the penstock leads to the effect shown in the video; during normal shutdowns, wicket gate closure rates are much lower and the water column doesn't reach the upper surge tank chamber).

Fortunately some field recorders allow dual recording which means that without external preamp you can simultaneously record the same input channel at two different levels (i.e. two ISO) while pre-setting a fixed dB offset of your choice (either among several given preset values or a freely settable value).
I'm also wondering about automatic gain settings and limiting. Many lower end field recorder only feature digital limiters, while good recorders feature analog (pre-ADC) limiters.

Are preamps with the very low noise floor you mention integrated in some recorders (SD, Nagra,...) or are still external top notch preamps required (and of course corresponding mics)?
Old 1 week ago
  #5
Lives for gear
 
tourtelot's Avatar
I record acoustic music so this certainly doesn't apply to any other genres but my boss once told me "the rule of 30." That is, to set the gain on your preamps, with condenser mics to 30dB. This is NOT scientific but it seems to hold true in my experience that 30dB is a pretty good place to start for acoustic music. Don't actually know why and I am okay to take the flames for this, but it seems to work.

Also, it's considered okay by the cognosci to record on a digital recorder at levels that do not approach 0DBFS, maybe peaking as low as -10, even -15 for scads of headroom. We just gain it up in post and since we add no noise, no foul.

But great mics and great preamps are important (although many good recordings have been made with mediocre gear.)

And yes, the preamps in Nagra and SD recorders are low noise and appropriate.

That's about ten cents worth of $.02s. (or -15 grains of salt)

D.
Old 1 week ago
  #6
Lives for gear
 

You need to know the SPL at the receiver mics reasonably well or be able to estimate it. Symphony orchestras can get to 125dB, chamber orchestras 108dB, string quartets 100-105dB at a main pair. Very roughly. An operatic soprano, anyone's guess. She can peel paint.

The feature I love about the Nagra VI almost more than its sonics, are the preamp pots calibrated in SPL dB. So once you know your mic sensitivity, and you have some idea of the SPL, then setting the pots is easy for digital full scale. Another wonderful ergonomic feature of the Nagra VI.
Old 1 week ago
  #7
Lives for gear
 
tourtelot's Avatar
[QUOTE=David Spearritt;13133417]You need to know the SPL at the receiver mics reasonably well or be able to estimate it. /QUOTE]

That's all good if you have any idea. Mostly, we don't. I'm doing a recording tomorrow where I don't have any idea. Solo tenor with five Baroque instruments
behind. No sound check/rehearsal. I just don't know. I'll let y'all know how I do.

D.
Old 1 week ago
  #8
Lives for gear
Quote:
Originally Posted by Schoeller View Post
(a rare turbine trip test initiated by generator load shedding at near full power, while such tests are performed from time to time they're kept to the strict minimum because they highly stress the equipment)
What is the purpose of the audio recording...what information will it convey ?
Is it to give prior warning of impending failure, or are you going to correlate SPL readings with generator load values. Is it to protect the hearing of humans down there ?

What sort of microphones are going to be involved : regular mics that we're accustomed to in this forum, hydrophones ?

Where will the mics be located...and will they be destroyed by SPL's or water during the exercise ?
Old 1 week ago
  #9
Lives for gear
 

Sorry to the OP for not reading carefully. So this is a turbine and not a string quartet. So are you recording this to check compliance with a statutory SPL requirement or are you requiring the recording for machine or condition monitoring analysis?

In either case you have 90dB to fit this recording into. Now this thing maybe loud, If you select your lower limit to be 40dB, then you can still record a signal without distortion (assuming mics are OK) to 130dB.

If its not loud or not that loud, plan on 30-120dB.

In the old days we had to fit such unknown recordings into 35dB on FM cassette recorders for vibration and 50dB for acoustic signals. We missed a few. These days it is luxury with a modern digital recorder.
Old 1 week ago
  #10
Lives for gear
 

[QUOTE=tourtelot;13133634]
Quote:
Originally Posted by David Spearritt View Post
You need to know the SPL at the receiver mics reasonably well or be able to estimate it. /QUOTE]

That's all good if you have any idea. Mostly, we don't. I'm doing a recording tomorrow where I don't have any idea. Solo tenor with five Baroque instruments
behind. No sound check/rehearsal. I just don't know. I'll let y'all know how I do.

D.
As our business is recording SPL, we should keep some idea in our heads.

Your phone will have some good apps to measure overall SPL and you could use this during rehearsal. I keep a few notes on groups in my familiar halls to assist when percussion or something comes along I can add adequate headroom to the typical levels. Singers are somewhat unpredictable.
Old 1 week ago
  #11
Lives for gear
 

Be sure to Record at a minimum of 24-bit.

It would be safer to record at lower level, rather than have clipping possible.
Old 1 week ago
  #12
Lives for gear
24 bit recording chains have amazing resilience
Just get it near the middle and they handle everything
Analog was very different and required gain planning and experience
I once recorded a long unrehearsed PTC from a presenter in front of an Avro Lancaster bomber during engine warm up
He was facing away from the Lanc and I tweaked his radio mic input down so the LEDs hardly glowed, half way through the piece the engines went to 100% power
4 Merlins are very noisy indeed, the 4060 and the Micron radio performed flawlessly
Recently I recorded a Gerontious performance with 2 x 40" bass drums, an amazing pulse of energy (with no rehearsal) the MD was convinced it would saturate but the NagraVI and MKH mics sailed through it
I watched some Eon Tusk re entry videos on Tube, shot by engineers, the sonic boom was rendered perfectly by mere video cameras, that is very loud indeed
Old 1 week ago
  #13
The dawn of the Turbine Quartet era cannot be far off...
Old 1 week ago
  #14
Lives for gear
 
tourtelot's Avatar
Well, as I suspected, no sound check and a pair of DPA 4011 in ORTF. Small Baroque ensemble, 6-7 pieces. Grand piano DSL, harpsichord DSL and really good counter tenor CS. Set my preamp gain to 35dB and was off by about 10dB which is a lot and probably disproves my theory. Easily fixed this morning in Pro Tools, but . . . .

Also put up a Schoeps Mk4 on a stick from a little underneath the vocalist and if I would have had a soundcheck, I would have added a touch. Big cathedral with flutter echo in two directions. Sheesh! Just an archival recording so no big whoop but boy would a sound check have helped. Oh well.

Oh, and I would have liked to audition a Faulkner array, two figure-of-eights, for a little more reach and maybe better room rejection but made the choice of the ORTF and went with it. Set up time was super short due to the morning service. The joys of live location recording, right?

The coolest thing was that the banjo player actually played the jaw of an ass.

Wild!

D.
Old 1 week ago
  #15
Lives for gear
 

Thread Starter
Thanks for all replies.


Quote:
Originally Posted by RJHollins View Post
Be sure to Record at a minimum of 24-bit.

It would be safer to record at lower level, rather than have clipping possible.
Is there any reason to record 16-bit if the recorder allows 24-bit without restriction? I expect that file sizes should no longer be an issue, i.e. a reason to revert to 16-bit. Of course I expect that internally 24-bit will be processed as 32-bit as processors handle data a 8, 16, 32, 64,... bits.

Most recorder restrictions I noticed are related to high sampling rates. Typically some recorders functionalities are not available (or only with some restrictions) above a sampling rate of 96 kHz, and/or the number of simultaneously channels to record is reduced.

I'm testing a TEAC TASCAM DR-701D which shows functional limitations at 192 kHz (compared to 96 kHz) but recording 16-bit or 24-bit doesn't imply any technical drawback.
As I use a 128 GB memory card (I posted the exact type in some topic), file sizes aren't a problem anyway. So basically unless power and/or storage would be an issue, there's no reason to not record at sampling rate of 96 kHz and 24-bit depth.

BTW I noticed that the noise floor (monitoring headphone output) slightly but audibly increases if sampling at 192 kHz instead of 96 kHz but I must check details.
Personally I set 96 kHz, I don't see any advantage sampling at 192 kHz and from a strict engineering's POV, 48 kHz is sort of quite close to 40 kHz = 2*20 kHz. Maybe 64 kHz would be a good compromise but it isn't usual so let's use 96 kHz.
Old 1 week ago
  #16
Lives for gear
Would it be possible to address some of the real world practical issues highlighted in post #8 ?
Old 1 week ago
  #17
Lives for gear
 
tourtelot's Avatar
Oh, do not go down that path. The "experts" will crawl out from under the baseboards.

It is my personal belief (read that again) that there is no benefit to recording at more than 96k, but that there is great benefit to recording at 24 bit and not 16 bit. FW all that IW.

I always record at 24 bit, even at 44.1 and it allows the correction of a lot of location recording sins. That and a gentle Waves L1 compressor on the bounce for final client files to get back gain without futzing with the dynamic range too much.

I think that "studio" projects have a lot more latitude to match dynamic range to DBFS than we do as location recordists, but I find that sample rate is not as important to what I do as bit rate. The bus driving by outside the church doesn't need 192k but the Izotope to take it out can sure use the 24 bits.

D.
Old 1 week ago
  #18
Lives for gear
 
tourtelot's Avatar
Quote:
Originally Posted by studer58 View Post
Would it be possible to address some of the real world practical issues highlighted in post #8 ?
What are YOUR answers to your question?

D.
Old 1 week ago
  #19
Lives for gear
 

Thread Starter
Answers follow.
Old 1 week ago
  #20
Lives for gear
 

Thread Starter
Thanks to everyone. My replies below are in no specific order.


Quote:
Originally Posted by studer58 View Post
What is the purpose of the audio recording...what information will it convey ?
Is it to give prior warning of impending failure, or are you going to correlate SPL readings with generator load values. Is it to protect the hearing of humans down there ?

What sort of microphones are going to be involved : regular mics that we're accustomed to in this forum, hydrophones ?

Where will the mics be located...and will they be destroyed by SPL's or water during the exercise ?
I only posted the video as example, it's not directly related to an application I'd be referring to.

Large rotating machines are usually equipped with condition monitoring systems which permanently measure and analyze various parameters.
Typically there are accelerometers and shaft position sensors as well as of course the usual bearing temperature sensors, etc.
Often there are dozens of instruments and if including generator temperature measurements there can be several hundreds (includig also other instrumentation). Nowadays data is nearly exclusively processed digitally, both be specialized hardware and generic programmable logic controllers (PLCs).

As the ("audible") audio spectrum is often too restricted, typically wider spectrums are recorded and analyzed (mostly by digital signal processing) for various purposes. Both installed permanently and using portable equipment. Very common is bearing monitoring.

That said, basic audio recording can be useful for comparisons using a reference recording though machinery is often difficult to record due to the noisy background and also maybe the same mic or at least the same model should be used.
Audio is used for example to detect leaks, there are special devices but now that I mention it I'll try once using a metal body mic (shotgun).


Quote:
Originally Posted by Coconuttopdrop View Post
If you don't know the dB's at source, I don't recon there's a proper way to do it
Evaluate your distance from the sound source, take into account room reverberation, (you can ask someone to pop a ballon at the same distance and check those levels).
Loved the videos!
Interested in hearing what other guy's have to say about this.
The video is sort of unique as in most cases surge chambers are not accessible (unless for inspection during shutdown), they're mostly cavities in the rock with vertical shaft and/or inclined shaft. The top of the shaft can often be accessed but it would require a very good lighting to film it, especially as the water column is not supposed to come too close to the top of the shaft (typical heights are often over 100 m (about 3333 ft)). The motion of the air and water will still make noise but it won't be the same as in the video.
Also there's really a lot of reverberation due to the rock walls (no idea how it would sound to record interviews).

Several posters mentioned that being able to at least roughly estimate the SPL would be priority.
Overall I expect that in many cases it will still be possible to very approximately estimate some maximum SPL, for example will it be some loud indutrial noise, or maybe a jet taking off at a distance of 200 m, or a detonation? In many cases I expect that experienced sound recordists will still have some idea.

The other main point was to keep plenty of headroom in order to avoid clipping.
If clipping occurs the audio becomes useless as it's for sure worse than some raised noise due to increased headroom.

If really high SPL are expected, maye one option would be to set levels simply based on the maximum mic output, fairly close to the maximum specified dynamic rate SPL (one mic I can test survives a max. 143 dB SPL and the dynamic range is 131 db SPL (sensitivity 25 mV at 94 dB SPL at 1 kHz)). Now I'd have to check what I get if I'd set the levels to get maybe -3 dB at 131 dB SPL (it's mostly theorical as I can't calibrate unless simulating the signal, I mean I could try to find a high SPL source and use the estimation as "ballpark" reference).


@ tourtelot:
Setting levels in order to increase headroom is a good idea, in real life the overall result will depend a lot on the quality of the preamps.
Peaking at -10 dB or even down to -15 dB requires correspondingly good preamps but obviously as said it's way better to be too low than risking clipping.


Referring to the already discussed sampling rate and bit depth, is it incorrect to simply go for 96 kHz / 24-bit if there are no specific reasons to use a lower sample rate and/or lower bit depth (the latter not being a recommended option as several posters mentioned it above)?
I simply mean that 96 kHz / 24-bit seems some sort of good compromise and as long as the recorder can handle it why not using that setting. I don't expect 192 kHz to be useful, especially as there are restrictions for the (half-)prosumer recorder I mentioned, also it requires more power.

Dual recording at two different levels could be interesting though it should be checked how it has been implemented.
What I'm not sure is about the limiter and/or automatic level control. Those I couldn't test yet and the problem is that those functions are digital (referring to the discussed recorder, thinks look better with analog limiters).
I've seen some YouTube videos showing the startup of very large diesel gensets (above 2000 kVA) where before the start you hear someone talking but once the engine is running the SPL must be very high but the recordings don't clip. I suppose automatic gain control was used in such cases (here referring to camera audio).

(This reply to be continued.)
Old 1 week ago
  #21
Lives for gear
 
Tommy-boy's Avatar
 

Quote:
Originally Posted by tourtelot View Post
I record acoustic music so this certainly doesn't apply to any other genres but my boss once told me "the rule of 30." That is, to set the gain on your preamps, with condenser mics to 30dB. This is NOT scientific but it seems to hold true in my experience that 30dB is a pretty good place to start for acoustic music. Don't actually know why and I am okay to take the flames for this, but it seems to work.

D.
Low 30s is usually my starting point too. I have some gigs where i dont get a proper sound check. I start here, estimate volume of group, mic distance, and guess at best range. I often adjust after first song. For critical recordings, i run a second recorder abt 10 decibels lower.
Old 1 week ago
  #22
Lives for gear
Quote:
Originally Posted by tourtelot View Post
What are YOUR answers to your question?

D.
My answer lies in the word rhetorical....which typifies the tone of the entire thread, while those 2 turbine videos in post #1 could lead one to think otherwise.

I was stupid enough to think we might be delving into the world of remote sensing data gathering methods...silly me
Old 1 week ago
  #23
Lives for gear
 
pkautzsch's Avatar
 

If it's un-repeatable, better use all options you can handle.

I'd go for at least a variety of different sensitivity mics. A typical location shotgun, maybe an SDC omni pair if we expect (and want to reproduce) a lot of LF content, and why not use a dynamic like MD21.
Dual gain for each mic, why not. We might end up using more than one 8-track recorder.
24 bits - yes. 144 dB of theoretical dynamic range is more than any electronic component (including ADCs) will yield because of thermal noise.

BTW that "moderator talking, then a loud engine starting up" thing mentioned eariler might not be AGC, but rather manual gain control or careful crossfading between differently gained mics.
Old 1 week ago
  #24
Lives for gear
 

Thread Starter
Quote:
Originally Posted by studer58 View Post
My answer lies in the word rhetorical....which typifies the tone of the entire thread, while those 2 turbine videos in post #1 could lead one to think otherwise.

I was stupid enough to think we might be delving into the world of remote sensing data gathering methods...silly me
Audio for pleasant HiFi sound and audio for technical purposes are different matters.

For such technical and scientific purposes you'd typically use something like this:
https://www.bksv.com/-/media/literat...215.ashx?la=en
or if link doesn't work:
https://www.bksv.com/en/products/dat...s-and-hardware

It's about industrial/lab test and measurement equipment with objectively defined specs to allow reproductible results, far away from audiophoolism.

Also as I mentioned it, depending on the application, frequencies of interest may cover a wider range than the common 20 Hz to 20 kHz. Used sensors would mostly be specialized calibrated technical microphones as well as other transducers.
BTW Brüel & Kjær manufactures a wide range of equipment used by mic manufacturers (also to acquire and record mic specs found in datasheets).

Indeed I'm wondering how audio recordings would sound if done using for example Brüel & Kjær test lab equipment instead of high-end pro audio equipment. Would it sound great? Or rather "sterile" because of the wanted technical accuracy? I've no idea as I never used Brüel & Kjær audio instruments excepted some desktop spectrum analyzer. Some time ago I used to know their condition monitoring products (vibration monitoring of rotating machinery) but IIRC they sold the division, now Brüel & Kjær Vibro known for the famous and widespread Vibrocontrol rackmount systems).

Data analysis is performed with various specialized tools.
For common applications a lot of more or less complex digital signal processing is handled in the background as the user is mostly interested to know if the monitord equipment behaves anormally, typically there are several thresholds, le last one initiating an emergency shutdown. Often channels are redundant to avoid spurious trips due to instrumentation and control system failures. In some cases a single spurious trip can cost more than an instrumentation rack (mostly due to production losses as some processes are expensive to restart).

That said, "HiFi" (i.e. here meant as "non-technical") audio is still a very special and extremely demanding, especially also due to the incredible dynamic range. For most common industrial measurements the required precision and resolution are fairly low.
I even suppose audio is the most demanding domain when it comes to widely used technology (i.e. consumer products), I mean in other common domains technical requirements are less demanding in terms if dynamic range. For example digital video requires very high data communication bandwidths but the dynamic range a such is much smaller (even with so-called high-dynamic-range video), general purpose industrial instrumentation often doesn't even require true full 8-bit resolution (even if most controllers obviously have higher resolution DACs).

There are many industrial data acquisition and recording systems, many are modular 19" rackmount designs, some modular systems are standardized, some are proprietary.
There are also rugged recorders as well as wireless solutions.
Old 1 week ago
  #25
Lives for gear
 

Thread Starter
Quote:
Originally Posted by pkautzsch View Post
If it's un-repeatable, better use all options you can handle.

I'd go for at least a variety of different sensitivity mics. A typical location shotgun, maybe an SDC omni pair if we expect (and want to reproduce) a lot of LF content, and why not use a dynamic like MD21.
Redundancy is the answer, though it can sometimes be tricky due to various constraints. For example being granted access shortly before having to record or unusual environmental or access conditions.


Quote:
Dual gain for each mic, why not. We might end up using more than one 8-track recorder.
24 bits - yes. 144 dB of theoretical dynamic range is more than any electronic component (including ADCs) will yield because of thermal noise.
As said, with modern recorders I don't see any valid reason to not record 24-bit unless there would be recorder-specific technical limitations requiring 16-bit in some cases.
Interestingly, power differences are most often not clearly detailed so it's impossible know how much power a given configuration will require until testing it (depends on many factors like number of used channels, sampling rate, display backlight intensity, supplied phantom power,...).


Quote:
BTW that "moderator talking, then a loud engine starting up" thing mentioned eariler might not be AGC, but rather manual gain control or careful crossfading between differently gained mics.
I was referring to an amateur YouTube video, probably shot with some DSLR mic with a windshield (and probably no separate recorder). I expected some sort of deadcat windshield due to the strong ventilation air flow, inside containerized gensets the cooling air flow is quite strong, for example at around 30 m3/s (about 64000 cfm but I didn't check the unit conversion) for a large genset).
If I can find the video I'll post the link but the audio sounds in no way professional but interestingly despite the very high SPL (maybe around 130 dB(A) or somewhat above) there's no clipping.
Old 1 week ago
  #26
Gear Nut
 
Haydn's Avatar
 

I sometimes use a Y split cable to send the output of one microphone to two pre-amps, one of which is a low gain safety channel.

edit: That Youtube video in the OP is awesome btw!

Last edited by Haydn; 1 week ago at 03:20 PM.. Reason: Edit
Old 1 week ago
  #27
Lives for gear
 

Thread Starter
For critically important work redundancy is the only option and, where possible, diversity.

To summarize there seems to be some general agreement that setting levels in order to prevent clipping is paramount, even if that means setting lower levels than usual. Also possibly dual recording with a fixed level offset can be an option.

Interestingly automatic gain compensation/level control as well as using limiters hasn't been brought up, so should it be concluded that it's better to set fixed levels with sufficient headroom rather than relying on some automatic variable level control and analog limitation (digital limiters are useless if the signal saturates the preamp)?


@ Haydn:
I don't know the person who recorded the video, he's an employee of the hydroelectric company. Unfortunately Kraftwerke Sarganserland AG is nearly totally owned by Axpo Holding which is a very bureaucratic company making special visits close to impossible (formerly there were lots of small independent hydropower plants operators and often you could visit plants, dams or so without problems).
It would be interesing to have a professional video made during a future test but I don't know how often such tests are done, also it's one of the extremely rare plants where the upper surge tank can be accessed outside maintenance shutdowns (the ramp near the stair is for the inspection vehicle of the inclined shaft, the overhead crane is used to inspect the vertical shaft, at the beginning of the video an employees can be seen looking down the vertical shaft, later he walks along the right rail of the overhead crane; both vertical and inclined shafts are interconnected but levels slightly differ due to dynamic effects; most surge chambers are simply cavities in the rock, sometimes there are only shafts without chambers; in most case there's no access during operation or you can just look down the shaft but as it's usually quite dark not much can be seen).

Here's a video of another rare periodic test, Hongrin (FMHL) operated by Hydro Exploitation SA (Alpiq group), as it's every 10 years maybe there will be a test next year:
YouTube

Couldn't find any video with good sound (it's quite loud but not dangerously loud). The Château de Chillon (Chateau de Chillon — Wikipedia) in the background is the most visited Swiss monument and just behind the castle was a former "top secret" underground military facility (Fort de Chillon — Wikipedia).
Indeed the water flow seen in the video is much lower than the flow during regular production or pumping operation but there's not much to see as intake and restitution happen under the level of the water (located just below the vane regulating the jet seen in the video), the motion of the water can be guessed in the background of the first pic of the link for the castle; the two power stations are underground in the mountain very approximately located under the highway bridge.
Old 1 week ago
  #28
Lives for gear
As this seems to have become the default "refuge for obsessives, polymaths and autodidacts" thread, I thought it the right place to throw in this Sample rate Converter Comparator from Infinite Wave Mastering in Canada...since the banter is losing momentum and needs a pot-stir

Preamble: "Originally developed in response to discussions amongst mastering engineers on Glenn Meadows' Mastering Webboard, our Sample Rate Conversion Comparison Project has grown to become a very popular resource.

With a database of over 200 different conversions, this portion of our site enables you to compare the performance of most software and hardware sample rate converters on the market"

SRC Comparisons

Just select a pair of converters from the drop-down menus, select test parameters (eg sweep, passband, 1KHz, phase, impulse, etc).... and then let them do their shoot-out. Remember this only tests 96 > 44.1 k....but that's useful for anyone here who records at that sample rate and then down-converts to produce test CD's for clients, a not atypical scenario perhaps ?

Sit back and either be horrified, obsess, or dismiss as unimportant.... as your personal bent directs you

These can help with interpretation : SRC Comparisons - FAQ

SRC Comparisons - Help and Information

Last edited by studer58; 1 week ago at 12:38 PM..
Old 1 week ago
  #29
Lives for gear
 
Tommy-boy's Avatar
 

Who are you calling an autodidact?

Old 1 week ago
  #30
Lives for gear
 

Thread Starter
Maybe me. At least I'm not a velodidact.


Quote:
Originally Posted by studer58 View Post
As this seems to have become the default "refuge for obsessives, polymaths and autodidacts" thread, I thought it the right place to throw in this Sample rate Converter Comparator from Infinite Wave Mastering in Canada...since the banter is losing momentum and needs a pot-stir

Preamble: "Originally developed in response to discussions amongst mastering engineers on Glenn Meadows' Mastering Webboard, our Sample Rate Conversion Comparison Project has grown to become a very popular resource.

With a database of over 200 different conversions, this portion of our site enables you to compare the performance of most software and hardware sample rate converters on the market"

SRC Comparisons

Just select a pair of converters from the drop-down menus, select test parameters (eg sweep, passband, 1KHz, phase, impulse, etc).... and then let them do their shoot-out. Remember this only tests 96 > 44.1 k....but that's useful for anyone here who records at that sample rate and then down-converts to produce test CD's for clients, a not atypical scenario perhaps ?

Sit back and either be horrified, obsess, or dismiss as unimportant.... as your personal bent directs you

These can help with interpretation : SRC Comparisons - FAQ

SRC Comparisons - Help and Information
Not sure if it's the topic you meant, there's a discussion about sample rate conversion in another topic.

Wherever possible I'd just record at 96 kHz 24-bit and only convert at the very end of post for the delivered files whatever specs the clients wants. Like for photography, keeping a lossless format during all editing (or edit lists) and just preparing the files in the format requested by the client but always keeping the original untouched files as well as the final lossless files (before any downgrading or any other further customer-specific conversion).

If the used software isn't lame there shouldn't be issues degrading material due to insufficient data processing precision, i.e. even after a large number of computations, rounding errors should not reach a level where they could get noticed.
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump