Oftentimes when dealing with phono preamps, line level and prosumer PC audio I/O, there is no way to adjust gain of the signal hitting the audio I/O device. @bdiament some years ago recommended here (I can't find the post for the life of me) that even as low as -18dBFS was more than enough signal level for an ADC capturing at 24 bits. -3dBFS is great if you can hit that without ever hitting 0dBFS. When I am able to adjust gain during recording, I usually shoot for -10dBFS to -6dBFS.
-18dBFS seems quite low, but I am not going to argue. It is better to error on the soft side than the loud side that I know from experience.
Anyone looking for a version of RX Advanced, can't go wrong with this deal. Save over 50 percent. http://www.ebay.com/itm/iZotope-RX-...mputer_Recording_Software&hash=item2ecda6fac4 Make sure to offer much less. Note: Here's the only requirement, Q: What kind of eligibilty is required? A friend of mine is student, which means any facility is going to be accepted? A: A student ID is all we need.
Don't worry, they will counter offer. $540 After you accept or they accept the offer. Send them a scan or photo of your wifes ID. If she doesn't have one, then call them to find out how to proceed. Takes them 1 day to process and email you the license code. Meanwhile go ahead and download the software online from the company's website. Note: They also offer the non Advanced version for students educators at half price.
I have been doing quite a few needledrops lately. If the records are in very good condition, I use manually declick very carefully in Adobe Audtion, only selectiong a few samples for each click. I want to make sure that I don't ruin any of the audio information. All records are professionally cleaned before recording with a high quality RCM. But some old records still have many mostly smaller clicks, and it is very time consuming, pretty much impossible for me, to manually declick these recordings. So, I downloaded Click Repair to test it. I want to make sure that the audio signal gets affected the least possible. What settings should I use to be on the safe side (regarding not harming the audio)? I have made some initial tests with shorter sample files, but I am not certain what settings to use for old records. My specific questions: a) which "DeClick" level should I use (I made some tests with 9 and 15)? b) should I use "DeCrackle" also, and if so, at which level and which method: simple or wavelet 2x/3x? c) should I leave "pitch protection" off? d) should I use "reverse"? e) should I use "automatic all (100)"? Since I am doing the processing on my computer and it is a bit more cumbersome to get the files to my main Stereo (for sound comparisons done by myself), I would appreciate a bit help with my learning curve by experienced users. Thanks a lot!
Yes, I'll give some answers. a) which "DeClick" level should I use (I made some tests with 9 and 15)? this is exactly where my use range is. b) should I use "DeCrackle" also, and if so, at which level and which method: simple or wavelet 2x/3x? Decrackle off / simple. c) should I leave "pitch protection" off? off d) should I use "reverse"? yes e) should I use "automatic all (100)"?
I have another question. I created a needledrop that has some annoying sibilance in the vocals (the "ss" thing). Is there some knd of method to reduce this, some kind of deesser / de-esser? I tried to use the "normal" setting in Adobe Audition, but that didn't help. Any advice would be very welcomed.
FWIW my preferred De-Esser is the one built into Tracer's Diamond Cut Audio Restoration software, current version is 8.5.
I've found the most effective way to deal with such sibilance is to convert the file to Mid/Side (MS) format, then select the sibilant only in the Side channel only (usually the Side channel ends up in what is normally the right channel. The selection also works better if you do it in spectral mode if you have one of the later Adobe Audition versions. I use iZotope RX Advanced), then apply a fairly aggressive declicking setting to the sibilant in the Side channel only. If the sibilant still remains, then try applying the declicking to the sibilant on the mid channel as well. I find this works better than de-essing because the latter tends to dull the treble transient more whereas a lot of sibilance from LPs comes from mistracking in the vertical direction, which ends up in the side channel info is read from LPs in a mid/side format and then matrixed into the traditional L/R. Once you're done, simply convert back to Left/Right mode. I believe Audition has presets in its channel mixer for this. While in M/S mode, you can even preview what the results sound like by choosing the convert to L/R in the channel mixer then clicking on the preview. This all may sound complicated, but it's really quite simple once you practice a few times.
I might add that you can also try de-essing the sibilant but only the Side channel for the same reasons I just mentioned. For declicking, I'm using RX Advanced with the Random click type and a strength of 7. I haven't used Audition in awhile but I don't believe its declicker has different modes, just strength, so play around with your settings until you get something that removes the sibilant but doesn't leave a mess. One other benefit of Mid/Side mode is that sibilants become much more apparent in the Side channel when you listen for them. You can really zero in on what's distortion and what's just part of the recording. For example, I have an old pressing of Surrealistic Pillow with a big pile of IGD damage on the song Comin' Back to Me. I was able to clean it up almost to perfection just by switching to Mid/Side then using the declicking method. It's a bit tedious to do a whole song with a lot of sibilants, but if you like the recording, the results are worth it.
Another thing to try with Izotope, that I have had good luck with, is if there is only a few "ss" sounds, you can highlight the "ss" sound and then use the mix to mono feature in the Channel Ops mode. I've found that will take away any harshness to the "S" sound and smooth it out nicely.
Do any of you guys use the azimuth adjustment feature in Izotope RX3? I'm wondering if it's necessary, I know that technically one channel of a record plays slightly ahead of the other because of how the stylus is slightly rotated from how the record was cut. But would it make a big difference? Also I could see it making a difference if perhaps the master tape wasn't aligned correctly when plating the record. I'm just curious. Also, if you use it, do you set it to "adaptive matching" or the suggest button? One way takes awhile, the other doesn't, which makes me think you'd get better results one way vs. the other. I can't find much information on the azimuth adjustment features anywhere online, practical applications especially. There was one time I did notice a big difference using it. I have A Certain Ratio "Shack Up" 45 which has a really poorly recorded live B-side. I was just messing around with it on various needledrops, and I used it on this track, and everything snapped into place. I'd never had the program move one channel as far as it did with this one, nearly 4.5 samples, or 0.2 ms. It sounded out of phase before using it, and the stereo image snapped into place after using it. It's the only time I've actually saved a track after using it. But I did use the feature to align a few things to mono as well. But once again, I'd like to know the difference between adaptive and suggest. Thanks in advance.
Thanks a lot for the input with regards to de-essing. This is all very new to me, I have never worked with M/S channels. I only have Adobe Audtion 3.0 without any extra plugins. I could upload a short sample of the issue, and maybe the people who are more proficient with that task could work on the sample and post back their result, and then I could try to replicate the results using Adobe Audtion. That would be much appreciated. Here's the link to a short sample: https://dl.dropboxusercontent.com/u/65618094/Uriah Heep - Salisbury (sample).wav
IIRC, Alexey Lukin from iZotope (also member here) mentioned once that the azimuth adjustment is not really designed for stereo material. I do use that part of the RX channel mixer for balancing channels when they seem a bit lopsided. Yes, of course one should use his/her ears, but sometimes a tool such as this really dials it in nicely and quickly.
M/S takes getting used to, but as I've written before, it's a tremendous tool for uncovering a lot of issues when working with needledrops. I have a shortcut key set up to toggle to and from M/S while editing/manual declicking/denoising, etc., and I find it indispensible. It's especially useful for unmasking or isolated those elusive thumps or clkicks that get buried in the mix.
Now there's a company I haven't heard of in years! Diamond Cut was the first audio software I ever used. But, the version I had back in 1998 was terrible! Anyway, my preferred de-esser is in Sony's Sound Forge 10.
Ooohh! Thanks for reminding me! Right now, I am working on the 45 mix of Donna Allen's "Serious", and it needs a bit of help. I have an azimuth correction plug-in that may help.
Would love to hear what all the different suggested tools for de-essing are capable of. The clip I posted is very short and should be good for such a demonstration... if you have the time and energy.
I have used the "Smooth" function in Sound Forge but don't know if Audition has a similar process. Also, listen to the section on each channel if it is a mono recording and copy or overwrite the best section to the other channel.
Ok, here is my quick take on your sample. I just cleaned up the hard ss's with Izotope declick function. I had the declicker, set for random clicks, on 10(the max you can go) and highlighted just the sibilant part of the word, it's not the best but for a quick fix, it didn't come out too bad. http://www.sendspace.com/file/w16z0n
And here's mine with RX. https://db.tt/iAQkBqNn Unfortunately, Audition 3 doesn't have the ability to select just certain frequencies in spectral view the way RX does. Perhaps with Audition 5.5 or 6 you could get better results.
I'm planning on running my stage into a Tortuga Audio LDR first, then to the ADC. Then I can keep the exact gain structure needed at the phono and adjust to maybe -10. Dynamic swings can be huge on vinyl. Great thread, good resource for me when I finally get back into some needledrop editing.