I challenge you to prove that HDX sounds better than TDM!

[RyanC]

Hey Mike the problem on that one was my mistake but we decided to try to take 8 out of the equation at that point. What we found out is that AAX plugins have either changed in the rewrite from TDM or the settings don't transfer 1:1.

Without plugins the files do audibly null. The reason for the hangup isn't under normal circumstances but what you find when you boost the null file by 96dB. And the problem is that what's going on down there in the test we did appears to indicate something wrong with the test method. IE you expect the file to either null or not, but you don't expect the chorus to null, and the bridge not to null and not to null especially on the right channel. Again this is all down around -120dBFS or lower (the music is about equal in volume with quantization noise).

We are still committed to figuring this out, but this is far more complex now because we have to determine what stage of the test may have caused this. I would ask that everyone respect that all 3 of us are doing this for free amid busy schedules and over multiple time zones.

If anyone is holding their breath, I will say that my personal interpretation of the data thus far is that if anyone can consistently pick HD native, TDM and HDX prints in a blind test, I will eat my shirt.

[UncleBubba]

I see we are drawing conclusions before the test is conclusive...fine...I'll draw my own conclusion for everyone...If a relevant comparison test is conducted properly between TDM and HDX and you can't hear the difference...you are in the wrong business!

How's that...does the work for bringing us closer to a good working test and results that can be analized...no?


right...no it doesn;t ...so how about we just cast those two conclusions aside and concentrate on doing a proper test...

so dudes...get PT 10!!!!!...Holy cow...and run the same mix with the same stock plugs (yes this is easy actually)...all AAX all DSP or a combo of DSP and Native...what is so hard about getting the free demo and doing this...obviously doing it from 8 has issues...just forget about why there...make it an even more fair fight for yourselves...use 10 guys...did I mention you should use 10?????

[RyanC]

Originally Posted by UncleBubba

I see we are drawing conclusions before the test is conclusive...fine...I'll draw my own conclusion for everyone...If a relevant comparison test is conducted properly between TDM and HDX and you can't hear the difference...you are in the wrong business!

How's that...does the work for bringing us closer to a good working test and results that can be analized...no?


right...no it doesn;t ...so how about we just cast those two conclusions aside and concentrate on doing a proper test...

so dudes...get PT 10!!!!!...Holy cow...and run the same mix with the same stock plugs (yes this is easy actually)...all AAX all DSP or a combo of DSP and Native...what is so hard about getting the free demo and doing this...obviously doing it from 8 has issues...just forget about why there...make it an even more fair fight for yourselves...use 10 guys...did I mention you should use 10?????

OK bubba let's not get nasty. Re-read my post please. I used 10 TDM and have already stated that the first snag is that there is a difference between stock digi AAX and TDM plugins themselves (could be just settings, could be they were re-written and sound different, that is a different test).

All I did state is that without plugins they null to greater than 100dB, and offered my perspective on that to those potentially on the fence, which I admit was a mistake.

Also your example highlights why doing this test isn't all that simple. Nobody stopped to think that maybe stock digi plugs got re-written into AAX and are now enough different to produce a (potentially) false negative.

[shelterr]

I just downloaded the PT10HD trial and I bounced a song out of 8.1 HD right before I upgraded and then bounced it again once I got 10HD up and running. There are convolution reverbs on the drums and vocals so these do not null BUT the beginning of the song is guitar only with a simple EQ on it and they DO NOT NULL. They line up perfectly visually but there is guitar sound to be heard down low. Also, my partner from across the hall picked the PT10HD bounce blind 3 times in my room. More open sounding and clear. Now I thought PT TDM was gonna sound the same due to the limitations of the process cards but it appears that something is going on. Also, I am loving the snappiness of PT10HD as well as clip gain, realtime fades (huge for beat detective), and RAM cache. I've got a few more weeks on my trial, but it looks like Avid is getting my money...

[Mike Caffrey]

Originally Posted by RyanC

Hey Mike the problem on that one was my mistake but we decided to try to take 8 out of the equation at that point. What we found out is that AAX plugins have either changed in the rewrite from TDM or the settings don't transfer 1:1.

Without plugins the files do audibly null. The reason for the hangup isn't under normal circumstances but what you find when you boost the null file by 96dB. And the problem is that what's going on down there in the test we did appears to indicate something wrong with the test method. IE you expect the file to either null or not, but you don't expect the chorus to null, and the bridge not to null and not to null especially on the right channel. Again this is all down around -120dBFS or lower (the music is about equal in volume with quantization noise).

We are still committed to figuring this out, but this is far more complex now because we have to determine what stage of the test may have caused this. I would ask that everyone respect that all 3 of us are doing this for free amid busy schedules and over multiple time zones.

If anyone is holding their breath, I will say that my personal interpretation of the data thus far is that if anyone can consistently pick HD native, TDM and HDX prints in a blind test, I will eat my shirt.

I know how hard it can be to figure out how to do a proper A/B. I once tried to A/B the Pendulum PL2 with the L2 and concluded that it's simply not possible.

It seems to me that there are two possible purposes to perform this comparison. One is to objectively evaluate Digi's claims that there is an actual difference and 10 sounds better. The other is to figure out if you will get better results with 10 and if they are enough better that it'sworththe upgrade.

I only cares about the latter, even if it's a placebo effect, as long as it's genuine it's worth it. If I get better mixes because I believe it sounds better and work longer until it does, or if I'm cheap and only use Digi plugins and the new ones are better, if the system is faster and I get 59 minute of mixing done in an hour instead of 50, if something about the way I think is better suited for clip gain than automation - I don't care what the explanation is, all I care is if the results are better.

My sense is that everyone involved I the test is more concerned about the objective approach. In the end, I don't think it's a problem to end up with a subjective or blurred conclusion as long as it's not presented as an objective one because the practical results simply matter more.

[RyanC]

Hey Shelterr- Well the original goal was to compare TDM to HDX not 8 to 10. What plugins are on there? My gut feeling is that 10 has really fixed some ADC issues and that's the biggie, but that is yet another test. All it takes is one plug on an aux that isn't set to %100 wet for this to be very important, and ADC on auxes was a mess in 8... How many dB does it go down? I'm going to double check what our 8 vs 10 was.

Mike- Yeah man I'm already sold on 10, but I know what you mean, instantiating a plugin without a dropout isn't just a cute feature at all, especially if you have a client in the room. I bought a HD|N and love it. Still there is nothing wrong with loving HD|N or HDX for the speed and how that improves your workflow instead of attributing it to a sound quality difference, if that is the case.

[shelterr]

The guitars in this test were not routed to any busses. The beginning is just one guitar left, followed by one guitar right each with one waves puigtech eq plug on them, both sent to outs 1-2. Gettin the waveforms to line up was easy and I'm not sure exactly how far it went down but it didn't seem high enough to indicate a problem with the test. I can post the guitar intro files for people to compare for themselves but it seems pretty conclusive.

[T_R_S]

Originally Posted by RyanC

What we found out is that AAX plugins have either changed in the rewrite from TDM or the settings don't transfer 1:1.

True since AAX DSP chips and TDM DSP are totally different processors.
There was an interview with Colin from McDSP and and another guy from Sonnox talking about the differences in AAX and TDM coding.
I've moved on from TDM all my systems are now gone and everything is HDX now.
TDM is dead end yesterday's technology, and at some point in the future HDX will be history too.

[shelterr]

Here ya go. Obviously once the music kicks in these will not null due to convolution reverbs but i included a bit of the intro so that people could just hear the difference. I'm not claiming to know why they sound different but these were done on the same computer, one in 8.1HD and the other in PT10HD right after i upgraded and the guitars in the intro have Waves Puigtech EQ's on them and are sent to outputs 1-2. No bussing or routing of any kind. I included a stereo bounce down as well as the dual mono files for loading into the DAW of your choice.

[UncleBubba]

So in the end...they SOUND different...even though everyone would LIKE them to sOUND the same...THEY DON'T...and the tree grew a little taller and the bush wider..

...oh and big surprise!...PT 8 SOUNDS different than PT 10...that is different versions of the SAME DAW folks...

[The Famous Yard]

PT8 doesn't have delay compensation and PT9 onwards does. That is probably the cause. It's not that simple (as this thread and the valiant testers are showing) to make sure the test is a proper one, i.e. only testing the DAW itself. There are many potential hiccups to take into consideration and other issues to eliminate. I really appreciate that some GS guys are endeavouring to do that test and eliminate those factors.

"If a relevant comparison test is conducted properly between TDM and HDX and you can't hear the difference...you are in the wrong business!"

@UncleBubba... If you say things like that before doing the test and getting your facts straight then I don't really know what to say to you. What good is a test if you don't pay attention to the results? To misquote you, if you can definitely hear a difference and there really isn't one, maybe you are in the wrong business.

[The Famous Yard]

At the risk of repeating myself, I am having trouble finding any place where AVID say their new software sounds better. The closest I've found is a short segment on an AVID vid on youtube where Vaughn Merrick says using 32 bit float gives you 1000db extra headroom. OK so you worry less about clipping, but I assuming we can manage not to clip 24 bit. He then says it sounds "much better, much cleaner".

I don't know what testing he has done (if any) or if that's just his impression without really testing, but I still feel as if AVID have put lots of "better sounding" phrases around their advertising PT10 but never make this claim. They do, however, allow others to make that claim, which I think is deceptive. Some of you guys have said they do say PT10 sounds better...Can anyone point it out? I am not having a go at anyone; I would really like to see what AVID say about their own software in terms of sound. I have so far found nothing.

I think hearing it from the horse's mouth would tell us all a lot more than some wild opinions being thrown around without backup.

[Roman Rowlands]

Blind test is the only true test. Some people don't like blind tests as they fear they will pick cheapest equipment lol x 2. That's what happened when Eric Valentine did a blind test between a top card and a much cheaper one.

Sent from my GT-I9100 using Gearslutz App

[shelterr]

Originally Posted by The Famous Yard

PT8 doesn't have delay compensation and PT9 onwards does. That is probably the cause. It's not that simple (as this thread and the valiant testers are showing) to make sure the test is a proper one, i.e. only testing the DAW itself. There are many potential hiccups to take into consideration and other issues to eliminate. I really appreciate that some GS guys are endeavouring to do that test and eliminate those factors.

If this comment is directed at the validity of the test I posted it should be noted that PT8HD TDM and PT10HD TDM have the exact same delay compensation. I am well aware of the variables that cause null tests to fail and the intro of this song has none of these variables. Same buffer setting, same delay compensation, 1 simple plugin on each guitar, no bussing or routing. The only difference is the version of the DAW. It should be noted that I fully expected the two to be identical given the limitations of the TDM cards.

[Roman Rowlands]

http://www.image-line.com/support/FL.../app_audio.htm

[The Famous Yard]

Originally Posted by shelterr

If this comment is directed at the validity of the test I posted it should be noted that PT8HD TDM and PT10HD TDM have the exact same delay compensation. I am well aware of the variables that cause null tests to fail and the intro of this song has none of these variables. Same buffer setting, same delay compensation, 1 simple plugin on each guitar, no bussing or routing. The only difference is the version of the DAW. It should be noted that I fully expected the two to be identical given the limitations of the TDM cards.

My bad. That is unexpected. We have nulled PT9 Native with PT9 HD, Ableton 8 and Logic 9 at our studio. If they are summing the same way it really suggests to me that there is no magic summing engine that gets things "more right". Fwiw bouncing Logic offline didn't null with Logic bounced in real time for some reason. Maybe automation doesn't come out the same. That would disturb me if I were a regular Logic user.

[stinkyfingers]

Originally Posted by The Famous Yard

...Fwiw bouncing Logic offline didn't null with Logic bounced in real time for some reason. Maybe automation doesn't come out the same. That would disturb me if I were a regular Logic user...

if you were a regular Logic user maybe your null test wouldn't have failed...

[passmore]

Originally Posted by The Famous Yard

PT8 doesn't have delay compensation and PT9 onwards does.

Sorry but PTHD has had ADC since version 6.4 introduced Feb 2003.

[The Famous Yard]

Originally Posted by stinkyfingers

if you were a regular Logic user maybe your null test wouldn't have failed...

I dare say you are right; I don't know Logic very well. My partner (who is a PTHD user and also doesn't know Logic very well) did that test. We were still quite surprised as it was Logic vs. Logic. Anyway this is OT.

Sorry about my ADC comment, said without thinking enough. Having done quite a few null tests, I kind of assume that if things don't null, something has gone wrong with the test. My prejudice!

[RyanC]

Originally Posted by shelterr

If this comment is directed at the validity of the test I posted it should be noted that PT8HD TDM and PT10HD TDM have the exact same delay compensation. I am well aware of the variables that cause null tests to fail and the intro of this song has none of these variables. Same buffer setting, same delay compensation, 1 simple plugin on each guitar, no bussing or routing. The only difference is the version of the DAW. It should be noted that I fully expected the two to be identical given the limitations of the TDM cards.

Who says the ADC between 8 and 10 are the same? 8 did not accurately comp delay on aux sends, 10 does. Also look a t your new sessions, more of the ADC is being done at auxes then at tracks like it was in 8. Personally with parallel stuff I have much fewer (no) problems with 8 then 10 in the ADC department.

It becomes yet another test, but from what I'm hearing/seeing ADC seems to have received an overhaul.

[shelterr]

Originally Posted by RyanC

Who says the ADC between 8 and 10 are the same? 8 did not accurately comp delay on aux sends, 10 does. Also look a t your new sessions, more of the ADC is being done at auxes then at tracks like it was in 8. Personally with parallel stuff I have much fewer (no) problems with 8 then 10 in the ADC department.

It becomes yet another test, but from what I'm hearing/seeing ADC seems to have received an overhaul.

The guitars at the beginning of the song are NOT sent through an Aux track. Waves Puigtech, out 1-2. And they align visually perfect so I'm quite sure ADC can not be to blame in this case.

[CeretoneAudio]

Originally Posted by shelterr

The guitars at the beginning of the song are NOT sent through an Aux track. Waves Puigtech, out 1-2. And they align visually perfect so I'm quite sure ADC can not be to blame in this case.

version 9 was the first version to implement pan laws (the stock pan laws between 8 and 9 are different) .. make sure the pan laws are the same

[UnderTow]

Originally Posted by shelterr

The guitars at the beginning of the song are NOT sent through an Aux track. Waves Puigtech, out 1-2. And they align visually perfect so I'm quite sure ADC can not be to blame in this case.

By visually, do you mean all the way down to the sample level?

Alistair

[shelterr]

Down to a sample level. However I never checked pan law settings. For the record, I never claimed to know why the PT10 bounce sounded better just that it did sound better and didn't null. If pan law is the reason, why was the default setting in 8 worse sounding and why didn't anyone tell me to change this years ago?!? This whole idea pretty much proves the ones and zeros are anything but, and that the sound quality of a DAW is in the hands of the programmers to a reasonable extent.

[mykhal c]

Originally Posted by shelterr

Down to a sample level. However I never checked pan law settings. For the record, I never claimed to know why the PT10 bounce sounded better just that it did sound better and didn't null. If pan law is the reason, why was the default setting in 8 worse sounding and why didn't anyone tell me to change this years ago?!? This whole idea pretty much proves the ones and zeros are anything but, and that the sound quality of a DAW is in the hands of the programmers to a reasonable extent.

to echo your last point, there is a good read on Samplitude Pro X in the SOS April issue where the creators, Herberger and Tost, talk about a mistake they made in coding in a previous Samplitude patch and it was brought to their attention by a user who 'heard' the mistake. the mistake, which they admitted they could not hear, was a dither omission on the 24th bit goin' from floating point thru the 24bit converter. their testing (search for the mistake), which they say took a long time, showed the user's hearing was right even tho they couldn't hear it. anyway, their testimonial seems to back your post...IMHO

[Amek Blue]

Hey, what Amek console did you have ?
I own I a Rembrandt / Galio and totally pleased with it.. for the money