Gear Review"My PCB Grinder Porter Grinder arrived in November. I’ve mastered several albums with it to get a feel for the box and learn its strengths and weaknesses. Yesterday my friend Ian Stewart and I took it over to Bob Katz’ place to do extensive testing on the unit, and today I am ready to present our findings."
Gear Review: GOLY Porter Grinder with Bob Katz
Authored by Head Engineer Matt Davis – November 2018
My PCB Grinder Porter Grinder arrived in November. I’ve mastered several albums with it to get a feel for the box and learn its strengths and weaknesses. Yesterday my friend Ian Stewart and I took it over to Bob Katz’ place to do extensive testing on the unit, and today I am ready to present our findings.
As an engineer who is relatively new to the world of hardware EQ ownership, I wanted to do due diligence in combating my own potential bias in assessing this EQ, and few things in this world are more objective than Spectraplus readouts and the opinion of Bob Katz about someone else’s EQ…so I will start with the cold truths and later address the aesthetics/sonics which are more in the realm of subjectivity. If you aren’t interested in the more technical stuff by all means skip ahead to our thoughts about the sound of the EQ in practice. These tests were conducted using a Crane Song HEDD Quantum D-A-D chain and Bob’s custom insert switcher. Furthermore, most of the amplitude response tests were conducted with use of a L-R transfer function to remove potential influence upon the results by the converters.
The noise floor of this unit is quite low for hardware with an internal power supply. As measured between 20-20k unweighted the unit measures -89 dBu/-84dBu or -107dBfs/-102dBfs with the +18 dBu calibration of the test converters. With A-Weighting this noise floor is -110dBu/-108dBu. The right channel’s slightly higher noise floor can most likely be attributed to the location of the Toroidal power supply transformer on the right side of the chassis, as the noise signature is mostly comprised of hum and induction related content at 120 and 240 hz. This is a very respectable noise floor by any standard, and when measuring an unweighted bandwidth of 96k the noise figures are roughly on par with the test converters (-80dBu/-98dBfs)!
The distortion figures for this unit are also vanishingly low. When fed by a +17dBu 1k sine tone the resultant THD is .003% THD. This is on par with Bob’s Maselec MEA-2 which also comes in at .005% THD with the same input. The harmonic distribution of the Porter is largely comprised of 3rd and a little 5th along with some 50/120/180hz hum on the order of -95dBu. An interesting discovery at this stage of the analysis was that when switched from L/R mode to M/S mode the amount of power supply/induction noise increases by a few dB, although still remaining at completely imperceptible levels.
Next we analyzed the front panel bypass of the unit. While the bypass does neutralize all bands, it is not a true hardware bypass as we discovered by analyzing the noise floor bypassed versus with the power disconnected. If you don’t use an insert switcher in your chain this may seem like a deal breaker, but the tolerance for neutral settings on this EQ are within a delta of 0.1dB (have a look at the Porter Flat graph) and the unit has no box tone that I, Ian, or Bob could ascertain at my studio or at Bob’s. I personally feel confident using it without a hardware switcher in my chain, but it is something that may be a consideration for others!
We next checked DC resistance at the input and output XLRs to analyze whether the unit is internally balanced. Pins 2-3 yielded a 51.18k resistance, with infinite resistance between pins 1-2 and 1-3. All output pin combinations showed infinite resistance, we highly suspected the unit to be internally balanced at this point.
Intermodulation Distortion was the next order of business and again the Porter came out on top. When fed 19k+20k tones a difference product of -85dBu was yielded at 1khz with an overall power level of -14. Just for reference, the intermodulation distortion of the HEDD Quantum converters was only 10 dBu lower than this EQ, and the MEA-2 was 3 dBu lower. Clean is an understatement here.
But how much headroom does it have? We actually had to recalibrate the D/A converters and test the EQ output with a scope in order to find the clipping point for this EQ, which (in L/R mode) comes in at a substantial +27dBu for 1% THD, and a measured .005% THD at +23dBu…at this point we assumed the internal topology to definitely be balanced. An interesting discovery here was that the clipping level for the Mid/Side board was quite a bit lower, clipping at +22.5dBu and a .01% THD at +21dBu. We suspect the gain staging of the Mid channel encode stage is the culprit here. This is a minor issue but one that should be taken into consideration when choosing converter trim settings if you intend to use the Mid/Side function on this EQ, I have adjusted my converter trims to +18dBu in light of this discovery.
As I mentioned earlier the neutral settings for this EQ are astonishingly accurate. We noticed a subtle roll off starting around 20 hz and landing at -0.4dB at 10hz, I’ve been in contact with Gustav who tells me this is within spec for the unit. Outside of this anomaly the amplitude response delta is on the order of 0.1dB across the audio spectrum. Furthermore, the gain values read very true, and the tolerance between left and right band gains is incredibly uniform on an order of less than 0.1dB for all the bands we tested, bravo Gustav! The tolerance on frequency values versus those measured was a bit larger. For instance a bell on the 1k band measured as having a 1100hz center frequency on Spectraplus (as shown in the bells graphs). This is something to be mindful of, though in my case certainly not a nuisance. These center frequencies (though differing from the label) were also exceptionally consistent between the left and right bands.
So what about these infamously tight porter Qs? They certainly are tighter than the widest Q on Bob’s MEA-2 (have a look at the Wide Q comparison graph), but surprisingly they also don’t get as tight as the tightest Q on the MEA either (check out the Porter wide vs narrow q graph). The resolution within this range, however, is exceptional. With 6 values on the midrange bands and 5 on the high/low bands (one replaced with the shelves). Speaking of shelves, we discovered that the shelves in this EQ are slightly resonant, yielding a Pultec-ish shape with a cut immediately preceding the boost. Note in the shelves graph that the center frequency is at the zero crossing after the cut and before the boost. You might also notice that the low shelf is oddly shaped, resulting from the EQ not maintaining linearity below 20hz. These shelves are also very steep for a mastering EQ, on paper none of this looked inspiring.
At this stage in the process Bob wanted to hear these worrying looking shelves so we tried some acoustic music which was impeccably mixed by Andrew Diaz. The first impression when engaging the insert for the Porter was a subtle and very pleasant dimensional enhancement to width and depth in the mix, we attributed this to distortion in the unit’s op amps. The bells are admittedly on the narrow side. This is a surgical EQ moreso than a broad strokes one, but a very musical EQ which tightly integrates into the program material and with a very usable Q range for mastering. In the cases where Bob reached for a Q6 bell on the MEA, we compared it to the Porter with the same settings. Where the Maselec created a warmer and slightly more veiled sound, the Porter gave a more transparent and slightly more clinical result. These are both incredibly good EQs and the fact that they were so evenly matched when one costs half the price of the other is impressive. Despite the look of them, the shelves were very pretty and sweet in a way that wasn’t quite pultec-ish in character but also not like the Maselec’s truer shelf shape, at this point Bob redacted his initial reservations.
I feel like from a design standpoint this implementation of the NetEQ topology is an exercise in simplistic perfection. During the 2 months waiting for it to be built I did a lot of research on the history of the refinement of this EQ design, which has been chronicled in Gearslutz and GroupDIY threads. There seems to be a common thread running through the exchanges on those forums that Barry Porter’s original design has a certain magic built into it that can be lost by overzealously tweaking the design to achieve particular goals. While Gustav has made a lot of enhancements to the original design and turned it into a machine better suited for mastering, I feel he also respected the limitations of the topology and the spirit of what made the original design so great. I think this has yielded an implementation of it that, while being less complex/full featured than some of the other offerings on the market, might very well be the best sounding of the lot because of that choice. The power supply is a simple internal toroidal, the op amps are common ICs instead of esoteric discrete choices, the PCB traces aren’t made of adamantium…and it sounds damn good for it at a great price.
As for my experience in the last 2 weeks mastering with this EQ, well it has been sublime. Aesthetically it’s one of the prettiest mastering pieces I’ve seen, none of the pictures online do it justice. The Blore-Edwards switches are nice and clicky and not so stiff as to slow down one’s maneuvering, but also are definitely solid military spec switches. This EQ can be so incredibly subtle and result in a sound that doesn’t feel EQ’ed even with large amounts of band gain. I was a little reticent to buy a clean EQ when there are great ITB options for that out there like Equilibrium and wondered if in some way I would be paying $3500 for something I could get out of a plugin, wrong. While this EQ is on the surgical side for mastering EQs and very much a “Clean EQ” I’ve been able to create broad strokes tonal shaping that would have been much less transparent with my other tools.
When I set out to buy my first mastering EQ I wanted an ergonomic machine that would be an everyday driver that I would feel comfortable using on most music until I add color option/character EQs to alternate between later. I wanted a machine with very little box tone but enough vibe to justify itself. The majority of the material coming into my studio has a surplus of low end/low midrange needing cleaned up so I definitely didn’t want to be using a warming box as my everyday driver EQ and I wanted it to maintain transients and not soften whatever you pass through the machine. With those stipulations in mind I went with this EQ, and I’ve gotta say it’s delivered all of those things beyond my expectations. The times I’ve reached for Equilibrium in the last 2 weeks for anything apart from surgical high pass filters or linear phase high shelves has been very few. I believe the Q range to be perfect for my purposes, though I find myself leaning on the widest Qs frequently.
Apart from the EQ itself, Gustav Grinderslev has been incredibly patient and diligent, and an absolute pleasure to buy gear from, I really can’t stress this part enough. I spent a little time digging under the hood of the EQ and it is immediately obvious that these EQs are made with love and exhaustively calibrated. If you are reticent to buy handmade mastering gear…for whatever reason, rest assured that this guy is the real deal and delivers a mastering grade product by any metric you would like to choose. I will definitely be back to buy more of his hardware.
- Date : October 29, 2018
- Category : Gear Review
Thanks are due to Bob Katz of Digital Domain and Ian Stewart of Flotown Mastering for assisting in this Gear Review, as well as to EQ Designer Gustav Grinderslev.