https://on.soundcloud.com/YZFnGDIfHbxmGfWVw7
https://on.soundcloud.com/YZFnGDIfHbxmGfWVw7
|| Pre-Production Stage: Introduction and what remains ||
Starting this off, I would just like you to know that I believe it's best for you to take your sweet time going through the processes gradually. A factor that really delayed my time for this project was how my microphone arrived 2-3 weeks into the schedule, which meant only the beginning of the preparation period was met. This pushed all of the stages' duration to the maximum, and considering that I was doing this part-time, it proved to be challenging. If you can, prepare your equipment and start learning how to navigate unfamiliar software you're planning to use early because these also take time.
Moving on, as I have stated in the previous blog, I will be looking for a microphone instead of earphones. At first, with my budget, I could just purchase a lavalier mic from a well-known brand. The finalists for me were the Fifine K053 and BOYA BY-M1. I would say having its own sound card really makes the Fifine K053 stand out and tempting to buy, additionally, you can’t just ignore BOYA’s positive reviews at the same time, but there’s an actual comparison available on YouTube by WebStylePress which will be hyperlinked in this section of the blog if you want to check it out and evaluate the two yourself. As I was listening to it, I could only describe it in one word, which is ‘kulob’ ( This is actually quite ironic considering how my output turned out ). There are most likely other factors that could’ve contributed to that but I didn’t want to risk it considering that I only have one shot at this.
Eventually, I came across TNW K18 which has both positive reviews and quite the clarity in my opinion, at least, better than the two I was considering and a smartphone mic. I purchased it and it definitely sounds decent for the price of 600 pesos. It apparently has its own built-in soundcard and pop filter along with considerably noticeable cardioid pickup and denoise effect which you can activate using a button, has echo and volume adjusters, and RGB light if you’re up for it. Moreover, you can plug it on various devices using its dual-part cable that hosts both USB and Type-C plugs.
via Lazada advertisement
Thus, most of the equipment used for this project remained as it was initially declared in the Assignment 1, except for the smartphone mic that wasn't used and the initial earphones replaced by a microphone. The final list of hardware and software tools used for this project are as follows:
Hardware:
TNW K18 USB desktop microphone (input and output medium)
Xiaomi earphones (in-ear monitoring)
Intel i3-7100 based Windows PC (audio processing)
Software:
Waveform Tracktion (DAW)
VirtualPiano (Interactive website for piano recording)
Consequently, my signal flow chart catered to these changes:
|| Production Stage: Journey and challenges ||
This section of the blog will mainly tackle my production journey starting from setting up the DAW to receive microphone inputs to different challenges encountered until the recording was completed. Moving forward, the various challenges are outlined and emphasized for convenient navigation.
Main setup with the USB Microphone at the center of the desk
Experimental setup for situation diagnosis (My good friend Santa's setup)
A.) No input heard
At first, I was enthusiastic because there was definitely an equipment upgrade for me. Not long after though, I tried using it to record an audio track using Messenger for another course’s activity and the sound ended up getting glitchy and would cut off at some parts, almost like I suddenly choked on my saliva or ran out of air while recording. I switched to using the microphone on my pc and it wasn’t there anymore so it’s probably some kind of device or compatibility issue. Next, I used it to record in AudioMass as part of my initial plan to import audio tracks on my chosen DAW but it didn’t go too well. The track dynamics didn’t sound pleasant at all and would distort at the high parts even though the mic’s volume is set to almost mute and I keep proper distance.
After that, I decided to ditch that idea and try recording through the DAW itself which is Waveform Tracktion. It sounded much much better than AudioMass, but there were times when the mic would suddenly not pick anything up, like it’s just suddenly not working anymore so I got anxious because of course, if it’s a mic problem then that would mean I wasted my money on this microphone. Here are the some of the steps I took sequentially as the issue continuously persisted.
1st occurrence: Restarted the software and re-plugged my in-ear monitoring earphones
2nd occurrence: Checked and clicked the built-in voice button in my earphones just in case it was muted
3rd occurrence: Adjust the volume from the mic itself (Non-intentional as I had accidentally turned the knob to its lowest)
Some additional solutions I found on the internet if it occurs or still persists on your end are ensuring that your mic is where the input is actually coming from through your settings and restarting your computer.
B.) Buzzing/Robotic/Static sound along with monitored and recorded voice
Next, I tried to record, but I eventually realized there’s a buzzing coming from my input, almost like a delayed noise version of my voice. Some prospective solutions that I did was lower the gain and increase buffer size so that my low-end CPU can handle the input better. From what I know, a lower buffer size can decrease the delay in your audio but it can be quite heavy on your CPU. For Waveform Tracktion, you can hover at the top most area of its project interface and you can view your CPU usage. This first solution offered little improvement so, I opted to check out ASIO as it is said to improve the clarity of your audio tracks by making the audio processing more straightforward.
After my experience with ASIO, I came to realize that it’s mainly used for low latency live monitoring but since I have a direct monitoring through my earphones, I’m not really maximizing or making use of its main or all the benefits it offers but rather it serves as an extra insurance for stability on my end since from my research, it deals with digital artifacts better than Windows’ drivers, which became evident as there were some glitchy parts that somehow just kept popping up during the track while I was using the default Windows' driver.
ASIO website and driver interface, and software
From the device list above, you just have to power on which of the available devices you'll be using. For me, I have my K18 microphone as both input and output but I don't have them enabled right now as I used it recently as my general PC system's input and output device. Even if some of Realtek's are enabled, I believe it doesn't really do anything as I don't have the actual device where it's from, Realtek (R) Audio enabled to function and implement anything on both my driver and DAW.
Going back to the buzzing problem. I tried matching my PC system's and DAW's sample rates as well since I noticed it was mismatched and apparently it can cause misinterpretation of data within the playback device and audio source but unfortunately that didn’t fix the problem. To check whether this problem exists on your end, just hover to the upward arrow on your Windows' taskbar, right-click the speaker icon and open sound settings. Then, scroll down until you see the input section, click device properties, and check out additional device properties. Under the advanced section, you'll be able to see your device's default sample rate and bit depth which you can use as a reference when setting up values in your chosen DAW and/or driver.
The next solution I tried is moving the USB plug of my mic from the front panel to the back because it might’ve been caused by some electrical interference or noise bleed from other components passing through that particular area or port. The dzzz dzzz sound is much less now but I keep hearing a constant beep from my earphones, I pulled the earphone plug a little backwards and I did notice the change when I did but I have a feeling it didn’t really fix the problem completely.
I’m really determined to make the initial raw audio decent as much as I can because I believe it’ll be smoother working on it later on than relying heavily on post-processing.
I learned that USB microphones really have in-borne issues but now that the hiss, dzzz dzzz, and glitchy sounds are reduced on my end, I feel that I can start recording now.
C.) The return of the bees
After a while, I went back to record and when I listened to the playback that robotic buzzy sound along with my voice is back, and I became absolutely frustrated because I haven’t changed anything from the last time I did and I would be lying if I said I just went on my merry way to research and consult how could this have happened and what I can further implement to fix it. The next thing I did is isolate monitoring to my mic alone, you just have to right-click the part where you can see the bouncing meter, then you’ll see live input monitoring and turn it off. That’s what fixed it for me this time, but I still maintained my other settings that I did before, just to narrow the possibilities of other relevant issues coming up. This allows for only one path active, avoiding any overlaps with the two routings when both Tracktion and your USB microphone live monitoring are active.
D.) Silent exported file
Another problem I encountered is when the exported file won’t produce any sound once I click play. I was actually so held up on this, I feared for my future because thinking about spending hours on a mix just for your exported WAV file to not work later on, absolute tragedy. Anyways, as I arrived at the solution for this, it made me realize just how dumb I am because all I had to do was switch my system’s playback device to my mic where I have my earphones on. I initially had set it up to my default system speaker, which is Realtek Audio because I have my microphone as both the input and exclusive output device for Tracktion so obviously, the general system's audio wouldn't flow towards the microphone where I am trying to listen to.
I tried to have my input as my USB mic and my output as the Realtek one before, but it was so distorted like a radio who can’t pick up a signal, it’s mostly just noise with a pitch which was absolutely jarring. I had that idea because when I listened through my mic before, it sounded glitchy but when I switched to Realtek, it sounded okay. But now, it’s a different story and it’s off the table.
E.) Low volume playback
Coming back to my setup, the output started to sound off again. I was ready to record when all of a sudden I heard the playback’s sound too low or has been quieted for some reason, even with the built-in provided samples. I was then stuck with another problem and it was utterly frustrating because after all the work I did for days trying to fix everything and make it satisfactory, it had to give me another issue to work with and I am no master in this so all I could do was feel the pressure and frustration. At this point, I’ve come to the conclusion that Waveform Tracktion is absolute rage bait (I'm sorry but this was my live honest review at the time, thank you Waveform Tracktion).
Skipping to the solution, I reinstalled Waveform several times until the problem subsided. I know it doesn’t sound like a real solution at all, but at this point, whatever works I don’t dare to question anymore. However, I noticed that at times, it would display some kind of warning related to my CPU usage and the Panic button for resetting the audio engine would blink, implying that the problem could be coming from my equipment.
THE ULTIMATUM
As I was about to record, after reinstalling everything, the initial problem I had, which was the seemingly buzzing sound along with my voice came back even if everything I did to fix it before was in place so as my ultimatum, I decided to record using a course mate's setup. I just felt like going back and forth with the problem that doesn’t seem to want to be fixed just isn't as efficient anymore and I had to diagnose my reality better.
What I noticed first was that the live monitoring of her microphone seems to have more noise than expected. I would say that the cheap microphone I underestimated due to my circumstances seems to be working competently and competitively enough. However, after some time, I realized that despite the advanced setup, the problem still persisted and it seems like that's where the limits of my equipment lie. Thanks to my friend, I was able to diagnose the situation better and accept the fact that I shouldn’t be expecting something that high of a quality with the equipment I had, but I thought I could somehow. Nevertheless, it was a humbling experience and it made me feel like my previous trials and triumphs all ended in naught, but I do believe that these approaches can function to fix some real issues, so if it helped and worked for you, I would highly appreciate it.
So, I would say my tip is to diagnose and recognize your limits early on because I have a feeling it’ll be much easier and efficient for you to accept and cater to your current inherent capacity.
Into the recording part, I only recorded for around 2-3 hours for all the main vocals. Prolonged periods of singing can strain your vocals and lead to output changes, so for me, it's better to finish it as soon as possible or allot separate sessions for vocal recording to allow rest and stay consistent with your vocal quality. What I found helpful but actually a bad habit to develop is singing backwards, which means you're singing but you’re sucking the air in, or at least trying to as much as possible. It's like you’re eating your notes/words where low audio inputs of air from singing or breathing is avoided as much as possible since my mic turns them into additional noise, of which I have plenty already.
Concluding the recording stage, the following tracks were produced:
3 vocal tracks ( Tanjiro no Uta, Nezuko no Uta, and Second voice parts )
1 piano track ( Instrumental )
|| Post-Production Stage: Mixing and Final Thoughts ||
After recording, I proceeded to mixing and experimented with presets. For this part, it really comes down to your preference and goals but what I liked the most and used for my audio tracks for this project is the 8-Band Equaliser: Piano, Compressor: Live Vocals, 1-Band Equaliser: Hiss Removal, and Natural reverb: Sparkly Keys. From my version of Waveform Tracktion, you can access the available types of presets through the left side and click the arrow icon beside each folder to enable the drop-down menu containing its various presets. Not only do these helped mask most of the inherent noise my vocal tracks has, it also aligns with my goal and preferences with how my audio tracks are being presented. Optimistically, it sounds like an ethereal lullaby, almost like a siren song, but if you're not feeling it, it probably sounds like a song recorded from an outdated device or a soulful bathroom concert.
I also tried other presets to try and improve it as close to its raw version but it didn’t really work since the bees are really hard at work (buzzing) even then. However, I would still suggest exploring the presets you have because you never know, what you might really need is just around the corner. You don’t have to decide immediately or limit yourself to what sounds good for you at the moment, explore what else does and narrow it to what aligns most with you.
After organizing my vocal tracks, I began on my piano track. I initially was going to use an actual keyboard but unfortunately, it doesn’t work anymore so I decided to use an online piano where you can use your keyboard as the keys then you can record for a maximum of 30 seconds for free users, and download the resulting piece as a WAV file. However, it can get really challenging to hit notes, especially those of three and above because you're using a desktop keyboard that doesn't have the structure of an actual piano/keyboard instrument. Your fingers can end up anywhere and everywhere. For the final touches, I just continuously listened to it to check for any more parts that I can adjust or fix.
To end this blog, all I can say is that when doing this project, you should at least pick something you’re actually passionate about doing. If you’re not really into anything audio related or none of the relevant valid performances you could do interests you, pick a topic or concept that you like and I’m sure it’ll make things more fulfilling and fun despite the challenges you face because you’ll be passionate enough to get through them and do your work justice.