Gearslutz (
-   Audio Transfers, Restoration and Archiving (
-   -   iZotope RX 7 - processing order (

larshoel 14th August 2020 05:40 PM

iZotope RX 7 - processing order
2 Attachment(s)
I'm currently archiving hundreds of field interviews conducted on microcassette. These are being transferred to 24 bit/96 kHz WAV archive files; the next job is to clean them up to make listening copies. I do this in iZotope RX 7 Standard.

Microcassette technology was band-limited from about 400 Hz to about 4 kHz. (Make sense since they were mostly used for voice recording.) So I make an initial pass with the EQ tool, with a hi-pass shelf set at 200 Hz and a lo-pass at 6 kHz. Then I run Voice De-noise.

My thinking is that by first filtering out the "missing" audio frequencies, I make the Voice De-noise tool's job "easier." But is this a valid assumption? I notice I can drag both ends of the curve in Learn mode to somewhat mimic the EQ. (see attachments)

Also, I use De-Hum (if necessary) before Voice De-noise, since I've learned that Voice De-noise doesn't do a very good job on hum. Does that make sense?

Thanks to anyone who's used RX 7 in a similar project. I'd love to hear about your workflow.

All best,

GYMusic 14th August 2020 11:49 PM

Welcome to GS Lars!

I do a good amount of RTR transferring and archiving. Your approach seems logical, but I must admit for me every reel requires something different and sometimes nothing at all (EQ, noise reduction, S removal, hiss removal, click removal, level changes). How does it sound? The first and most important thing is getting a good transfer. With all the tools available to us (like Rx) we can experiment endlessly and if you don't like it, you can undo or go back to the original file. Even years later. Again, how does it sound TO YOU? Good luck.

gunofason 15th August 2020 01:54 PM

Hi Lars! As someone who has digitized hundreds of both mini & micro cassettes, there's an reasonable expectation of limited fidelity with these, so I wouldn't spend much time sweating over noise reduction for the entire collection (unless you enjoy doing it of course!). I use RX7 Advanced myself, but never have been fully satisfied with any of the auto or single-action dynamic plugins. For real effective results, it usually requires some experimenting and adjusting for each tape (and for each side of each tape).

Being hand-held devices, micro cassettes typically have some steady, narrow-band, electro-mechanical noise which can be removed most effectively with a single-band dehum at each precise frequency. There are harmonics of these noises, but it's best to do them one by one if possible.

Keep in mind, some devices had low-voltage detection and some didn't, meaning, if the batteries got low, the tape recorder got slower and slower. When you transfer back with new batteries (or a hardwired AC plug, which is highly recommended), it plays back faster and faster! This effectively shifts all of your eq/hum/noise problems up several frequencies. This can all happen over the coarse of a single side of tape, which is why noise reduction and eq efforts on micro cassettes are sometimes done in vain.

If you have the drive space, consider saving both the original transfer (Preservation Master) and then make your noise reduction edits to a separate file (Modified Master). This way, you'll always have the unmodified, original source copy to work with down the road when some amazing new plugin comes out! Most micro & mini cassettes are mono, so transferring and saving an extra file doesn't take a huge data footprint.

If this is for a client, and you do modify the original, consider documenting the technique you used and enter the data into an inventory spreadsheet or the CodingHistory of the BWF BEXT chunk. Your archivist will love you!

timtape 16th August 2020 11:50 AM

Hi Lars, already alluded to but the most important step in the signal chain is the playback of the cassettes. This cant be emphasised enough. Audio lost here often cant be fully recovered later. Noise and distortion added by a faulty playback machine such as amplifier hiss, hum etc may at best be only partly removed later.

Do you understand azimuth with tapes? For best playback quality especially in the high frequencies the playback head needs to be "tuned" to the tape being played, ideally on each side of each tape, even within one tape side. If not, the audio playback can sound far more muffled and unclear than it actually is on the tape.

Partly due to their relatively wide track compared to their very slow tape speed which is either half or a quarter standard cassette speed, microcassettes are extremely vulnerable to azimuth misalignment.

Adjusting head azimuth involves locating and then carefully turning the azimuth screw with a suitable non magnetised screwdriver for maximum program treble on each tape.

It sometimes seems like everyone has audio restoration tools but not too many know how to play back the original tapes with maximum fidelity. Audio restoration tools cannot bring back audio which was never captured in the first place.

I agree with you that especially low pass filtering is often important as there can be noise in the playback (often playback amplifier noise) much higher in frequency than the wanted program on the tape. Filtering this out in post seems to me a no brainer but as mentioned it can vary depending on the recording, especially with the tape speed. Judgement is often needed.

Hope this helps.