The tray comes out of blackout with a forest in one corner and a bald patch in the other. You can cut it, but selling it is hard. Heights differ, part of the tray has not arrived yet, part has already gone past, and the cause of that was set in the first hour after sowing rather than in the days under the lamp.
Quick glossary: Soaking - holding seed in water before sowing, from 2 to 24 hours depending on the crop. Priming - controlled pre-sowing hydration of seed followed by drying back; soaking is its simplest version. Weighting - a load or an inverted tray on the sowing during the first days, holding the seed in contact with the mat. Photoblastic seed - seed that needs light to germinate and stalls in full darkness.
An Even Carpet Is Decided Before the Lid, Not Under the Lamp
Across our own 70 crop cards, uneven or sparse emergence appears among the leading mistakes on 23 cards, 28 times in total. Not one of those mentions concerns lighting after blackout. Every one of them is about seed, water and weighting, which means the first day.
That sets the order of diagnosis. When a tray comes out uneven, there is no point checking the lamp or the schedule: the mistake finished its work long before either became visible.
The causes on the cards sort into four piles: soaking done differently from what the crop needs, water shifting the seed after sowing, weighting that is weak or uneven, and a seed lot or temperature that is wrong. In that order.
Soaking: Half the Complaints Live Here, in Both Directions
Soaking is required for 15 crops out of 70: pea, sunflower, fava bean, mung bean, chickpea, lentil, lupin, corn, oat, barley, wheatgrass, buckwheat, coriander, fenugreek, milk thistle. All of them are large seed with a hard coat. For the other 55 it is either unnecessary or harmful.
Skipping it produces ragged emergence. The cards say so directly: mung bean without soaking gives "sparse uneven emergence, part of the seed never wakes up," fenugreek gives "ragged emergence, 50% of the seed fails to germinate," fava bean gives "some seedlings already 5 cm, others not yet awake." The durations differ too: mung bean 10-12 hours in cold water, coriander 8-12, beet 12-24 with a rinse, milk thistle 18-24.
The reverse mistake costs the whole batch. Basils and flax must not be soaked: the seed coats itself in mucilage and clumps into gel, and the sowing comes out in lumps. Both cards record the same fix: sow dry only.
Over-soaking harms even the crops that need it. In a wheat trial, germination rose up to 12 hours of soaking and fell after that: at 20 hours it came in lower than unsoaked dry seed. Water volume mattered as well, with half and equal volume relative to seed weight beating double volume, where the seed suffocates in the excess.
Water and Weighting: Two Mechanical Causes
Irrigation after sowing moves the seed physically. Three cards describe it the same way: garden cress gets "seed movement after wetting, islands of density," ice plant gets "seed washed into piles by a strong stream," and oregano gets the same from a spray bottle. The fix on all three is one thing: wet the substrate before sowing, and after it use only mist or bottom watering.
Weighting appears on 52 of the 70 cards. It is there for contact rather than for evenness of the sowing: a seed lying on the mat with no pressure dries on one side and starts later than its neighbor. The radish card says "1 kg minimum, an even load," and borago says "increase to 2 kg and check that the weight is spread evenly across the sides."
The word "even" is what matters: a load sitting crooked gives a dense edge and a sparse opposite one. The remaining 18 crops carry no weight at all, being mucilaginous or slow aromatics for which pressure does damage, and their evenness rests entirely on precision in distributing the seed.
When It Is Not Technique: Lot, Temperature, Light
An old lot is not rescued by anything above. The corn salad card states it plainly: seed older than a year gives 30-50% germination and a sparse carpet. The fix recorded there is to buy seed no older than a year and calculate the rate from actual germination rather than from the figure on the packet.
Temperature works through speed. Chervil demands 16-18 °C, and above 22 it gives "sparse uneven emergence" along with stretching. The further a crop sits from its own window, the wider the spread between individual seeds in time to germinate.
Purslane, lettuce and tarragon are a separate case. Their seed is photoblastic: in full darkness it does not germinate, so a standard blackout gives not stretched seedlings but uneven emergence or none at all. All three cards state the regime separately.
Three Mistakes That Cost the Most
First, soaking out of habit instead of by the card. "Soaking never hurts" is false for 55 crops out of 70, and for basils and flax it costs the entire batch.
Second, expecting the tray to even itself out under the lamp. It will not: the gap set in the starting days only widens, because whatever emerged first shades whatever came up later.
Third, adding water when you see a bald patch. A bald patch is either shifted seed or a part of the lot that never came up. Extra water raises nothing there. What it does produce is local waterlogging and mold exactly where the carpet is thinnest.
How to Know the Carpet Is Even
Check it as the tray leaves blackout, no earlier and no later. At that moment the carpet should be continuous, with no exposed substrate and no lumps, and the height should look level across the whole area.
Count the seeds inside a 1×1 cm window in two or three places on the tray. If one spot carries half again as many as another, the problem is distribution.
If technique holds for three batches running and emergence is still ragged, test the lot instead of the tray. The gap between stated and actual germination closes in a few days with a damp paper test.
If you want to go deeper: Choosing Seed for Microgreens - the supplier and the age of the lot set the evenness you can work toward at all. What to do with a tray that came up evenly and still will not gain weight is covered separately: Stunted Trays.